Charting And Plotting

[aisobran]

I pushed some charting code up as plotAcc.py. It will take all the test accuracy results and write them to charts. I think bar charts make a little more sense than the line plots as each accuracy is separated by team. I do want to use the line plots to make comparisons on continuous variables. I decided to remove the train accuracy because I don't think it adds to the chart.

I changed the distribution name to tendency in the chart. I think that makes more sense as it's the tendency of the team to either run or pass.

I think we should have a scoring in place to communicate these results in a better way. I think the score should be:
Score = (Model Test Accuracy - Team Tendency) * 100

Team tendency = max(E[play=pass], E[play=run])

This tells us how many percentage points better the model performs at predicting run or pass than the team's actual tendency to choose run or pass.

[aisobran]

Also lines 6, 19, 31 should be altered to match your chosen file names and file structure.

[aadithya93]

Even I have pushed a code svmPlot.py and the plots are stored in the svmResults folder. I am trying to plot mean over the years per team now. With that my analysis will be done.